Modelling Extended Lactations in Polish Holstein-Friesian Cows.

Agnieszka Otwinowska-Mindur,Ewa Ptak,Joanna Makulska,Olga Jarnecka

doi:10.3390/ani11082176

Agnieszka Otwinowska-Mindur, Ewa Ptak + Show 2 more

Open Access

https://doi.org/10.3390/ani11082176

Copy DOI

Abstract

Simple SummaryMathematical models of lactation curves are functions that describe milk production on each day of lactation. These models are able to predict milk yields as well as provide valuable information applicable in breeding and management decisions. The aim of the present study was to examine different shapes of lactation curves for milk traits (i.e., milk, fat, protein and lactose yields and urea content in milk) modelled by the Wilmink function and by linear or squared functions between 306 and 400 days in milk (DIM). The results suggested that the course of extended lactations could be modelled by a nonlinear model, for example, the Wilmink function, for up to 305 DIM, and the linear or squared function could be more appropriate afterwards.The objectives of this study were (1) to examine different shapes of lactation curves for milk, fat, protein and lactose yields and urea content in milk fitted by the Wilmink function using the test-day (TD) records and (2) to find the function that best describes test-day records beyond 305 days in milk (DIM) for Polish Holstein–Friesian cows. The data were 6,955,768 TD records from the 702,830 first six lactations of 284,193 Polish Holstein–Friesian cows. Cows calved in 19,102 herds between 2001 and 2018. The following functions were fitted to TD data from each lactation: (1) Wilmink model fitted to the whole lactation, (2) Wilmink model fitted to TD records from 5 to 305 DIM and linear function fitted to TD records from 306 to 400 DIM, (3) Wilmink model fitted to TD records from 5 to 305 DIM and squared function fitted to TD records from 306 to 400 DIM. The present study showed that urea content in milk was modelled slightly worse than other milk traits. The results suggested that the course of lactation could be successfully modelled by a nonlinear model, for example, the Wilmink function, for up to 305 DIM, and by the linear or squared function afterwards.

Highlights

Mathematical models of lactation curves are functions that describe milk production on each day of lactation
The WIL305 model was slightly better than the Wilmink [11] model (WIL) function, which could be explained by the fact that beyond 305 days in milk (DIM) few TD data were available, so the fit of the WIL function was less accurate in those days
The values of the Q criterion were the lowest when milk and lactose yields were modelled by the WIL305 function, showing the best compatibility between true and estimated yields for those two traits

Summary

Introduction

Mathematical models of lactation curves are functions that describe milk production on each day of lactation. Mathematical modelling has been often used in the analyses of test-day data in dairy cattle [2,3,4,5,6,7,8] The usefulness of this method depends on how well the functions can mimic the biological process of milk production [3], so selection of the lactation curve model is associated with balancing between the fitting properties and the requirements for biological interpretations [9]. Other options are continuously increasing or continuously decreasing curves with no lactation peak or curves characterized by the occurrence of an additional peak later in the course of lactation The latter shape may be associated with seasonal effects and may characterize the cows calving in autumn in pasture-based farming systems [1]

Objectives

Methods

Results

Discussion

Conclusion